Donald Trump’s Usage of Classic Propaganda Techniques on Truth Social During the 2024 Presidential Election

Abstract

Guided by the ideational theory of populism and Jowett and O’Donnell’s framework for propaganda analysis, this study examines Donald Trump’s usage of propagandistic language and emotional appeals. The study classifies his posts on Truth Social during the 2024 American presidential election into the seven categories of propaganda tactics outlined by the Institute of Propaganda Analysis. These techniques include name calling, glittering generalities, transfer methods, testimonials, plain folk appeals, card stacking and bandwagoning. A deductive content analysis was performed on a corpus of the text of 1319 Truth Social posts (“truths”) posted by Donald Trump during the election, excluding image- and video-based posts. This study found a vast majority contain at least one classic propaganda technique. A contribution of this study is that it determines not only which tactics he uses as a case study of a populist communicator but who he targets and with which tactics. The study found five main targets: political opponents, political allies, immigrants, institutions, and foreign leaders. Findings demonstrate Donald Trump does not use propaganda techniques monolithically but rather as dependent on his target and with distinct patterns, to produce and reinforce narratives about in- and out-group messaging.

1. Introduction

“Matt Gaetz is a Great Representative for the Amazing People of Florida’s 1st Congressional District! Matt is a Warrior for our MAGA Agenda, who tirelessly works to Drain the Swamp, Secure the Border, Support our Brave Veterans and Law Enforcement, Defend our always under siege Second Amendment, Stand Up to the Woke Mob, and Fight the Never-Ending Witch Hunts from the Radical Left that are destroying our Country! Matt Gaetz has my Complete and Total Endorsement!”

This endorsement of Matt Gaetz by Donald Trump contains almost every tactic of propaganda as outlined by the now non-operational American Institute of Propaganda Analysis (Lee & Lee, 1939). It engages in name calling, referring to his opponents as a dangerous mob, and making glittering generalities of vague campaign promises with no actual substance (e.g., draining the swamp, making America great again, etc.). By transferring positive associations of heroic veterans to the Republican party, it invites readers to favor Gaetz while simultaneously transferring negative images of mobs and witch-hunts to liberal opponents. Through these propaganda techniques and show of his own support, Trump is attempting to gain favor for Gaetz’s candidacy. He invites potential followers to “jump on the bandwagon” led by a populist leader and metaphorically “stacks the cards” in the candidate’s favor, hiding any misdeeds or failures while highlighting or overemphasizing his successes.

This example demonstrates how Donald Trump’s rhetoric aligns with characteristics commonly associated with populist communication that is positioned both against perceived elite enemies and engaging with the common people. This study will further examine Trump’s use of propaganda through a lens of viewing him as a populist leader. Moffitt (2016) states that Donald Trump has been argued to have reshaped American conservatism; Rahn (2019) supports this, stating that when Trump created a deep distrust in the American system with his “impolite language and repetition of popular conspiracy theories, for example, or emotional arguments about immigration…” and claims that “democracy was being subverted by political and economic elites, which encouraged ordinary citizens to band together to reclaim control of the political system…” (p. 1), he was reelected for a second term in office.

It is important to begin by providing definitions for propaganda and populism. Benkler et al. (2018) define propaganda as a form of “communication designed to manipulate a target population by affecting its beliefs, attitudes, or preferences…to obtain behavior compliant with political goals of the propagandist” (p. 29). This study uses this as a foundational definition for studying propaganda found on social media, which can act as a tool for “empowerment and social change” or as manipulative “incubators of ‘fake news’ and propaganda” (Guess & Lyons, 2020). Populism is described as an ideology that coincides with other ideologies in supporting the common people while defying a corrupt elite (Mudde & Kaltwasser, 2017).

These definitions will serve as the operationalization for propaganda and populism, respectively, in the present discussion of how Donald Trump, a key populist communicator, crafts propaganda in his online Truth Social posts. This study seeks to explore how specific propaganda techniques are deployed differently across a range of different targets instead of considering propaganda a uniform communication strategy and to offer insights into the strategic logic of far-right populist communicators online.

2. Literature Review

As a populist communicator, Donald Trump’s rhetorical style combines elements of populism and propaganda. Therefore, in order to analyze Trump’s personal usage of these elements, a thorough understanding of both concepts and how they connect is essential.

2.1. Populism

The scholarly debate around populism centers on two main points: what populism is and if it even exists as we understand it in relation to other systems (Mudde & Kaltwasser, 2017; Mondon, 2022). Mondon (2022) warns if our understanding of populism is that which “obscure[s], deflect[s] and divert[s] attention away from processes of power formation and consolidation,” then populism and the study thereof is only “a simulacrum, a con” (p. 2). If populism is used disparagingly to mock, degrade, or deflect unwanted information through loaded terms such as “woke” (e.g., VanDreew et al., 2025) or “fake news” (Wardle & Derakhshan, 2017), etc., then, like Mondon’s point, it is an ineffective descriptor for academic study, even if such dismissal, distortion, or distraction are common practice in propagandistic messaging (Nimmo, n.d.).

Populism is not, however, solely a right-wing concept or ideology. Cooper and Avery (2021) explain populism is supported by both sides of the political spectrum, but that the political left and right understand the term “the elite” differently. Cooper and Avery (2021) state that progressives see the “infamous 1% who control the wealth of the country” as the enemy. Simultaneously, populist conservatives see elites as “fake news” media companies that mislead the people, consider the “deep state” as conspiring against the people, and position immigrants as propping up political elites, who benefit from their work at the expense of the middle class (Cooper & Avery, 2021, pp. 320–324).

Rahn (2019) describes populist attitudes as social, latent, and dispositional demands that must be triggered through specific contextual framing. As other scholars of populism (e.g., Mudde, 2004; Mudde & Kaltwasser, 2017; Moffitt, 2020; Rahn, 2019; Hawkins et al., 2018, among others) have noted, a universal definition is lacking, and as a “buzzword of the year… it is very often poorly defined and wrongly used” in both popular and academic discussions (Moffitt, 2020, p. 2). As the term becomes more widely used (and misused), it is not only important to understand what populism is but what it does (Mondon, 2022; Mudde, 2016). As such, Mudde (2016) argues the prior literature has focused too much on debating the correct terminology and definitions and only recently is asking how populist radical right parties and actors are relevant in today’s political communication. Therefore, the current study touches on the definitional debate before turning focus to differing perspectives on how populist leaders leverage the concept into their propaganda.

First, Mudde and Kaltwasser (2017) describe the common understanding of populism as a “folkloric” style of politics employed to “mobilize the masses” with “amateurish or unprofessional political behavior in order to maximize media attention and popular support” (p. 4). This behavior by populist leaders is understood as showing followers they are different from, and will stand up against, the current elite (p. 4), as demonstrated by the common praise of Trump’s status as a businessman rather than a politician. In this way, populism is understood as a strategic mode of organization to gain and maintain power (Moffitt, 2020).

On the other hand, the discursive-performative approach as outlined by Laclau (2005) describes left populism as a mode of antagonistic discourse between the people and elites (Moffitt, 2020). It introduces radical democratization as the solution to problems inherent to liberal democracies, therein mobilizing previously excluded and marginal sectors of society by bringing them into the conversation and changing the status quo (Mudde & Kaltwasser, 2017, p. 79). In doing so, populism achieves its main task in politics by constructing a coherent “people” and popular identity to rally behind (Laclau, 2005) while opposing the elite group perceived as a threat. However, Gorup (2021) argues that this approach is often reduced to “who the people are” rather than how or by what practices that construction occurs. Furthermore, Mudde and Kaltwasser (2017) frame “the people” simply as an “empty signifier” used to appeal by “generat[ing] a shared identity between different groups and facilitate[ing] support for a common cause” (p. 9). Other scholars (e.g., Jäger & Borriello, 2020) argue “the people” must somehow be “constructed and molded” by a strong leader rather than mobilizing themselves and could take the form of the “mob” influenced or manipulated by “demagogic superiors” (p. 743).

Jäger and Borriello (2020) also point out Laclau’s left-populism popular agency model may not adequately capture the nuances of today’s popular agency. Gorup (2021) thus responds to Laclau by outlining populism’s paradox of popular agency, which poses the question of how a group can even mobilize when it “does not know what it wants, because it seldom knows what is good for it” (Rousseau, 1968, p. 83, as cited in Gorup, 2021). While Laclau does speak of the importance of charismatic leaders in populist politics, Gorup (2021) argues this has not always been emphasized in interpretations of their work. Gorup (2021) believes they neglect the associational bonds that tie people together, arguing that people lack the capacity to act on a common purpose that necessitates the support of an intervening power to overcome the problem. Mondon (2022) echoes this, calling for further work on the construction of “the people” and “the elite” within the context of politicized democracy and reactionary politics.

Mudde and Kaltwasser (2017) further describe this as too narrow of an interpretation, considering populism to be a “positive force for the mobilization of the (common) people and for the development of a communitarian model of democracy” (p. 3). But they argue approaches that reduce populism to a political strategy used by strong, charismatic populist leaders governing “on direct and unmediated support from their followers” to amass power are unsustainable, as leaders eventually die and must be replaced (p. 4). The ideational theory of populism was widely popularized into a dominant theory and approach in the study of populism by Mudde and Kaltwasser (Mudde, 2004, 2016, 2017; Mudde & Kaltwasser, 2013, 2017).

While many approaches describe populism as a “discourse, an ideology, or a worldview” (Mudde & Kaltwasser, 2017, p. 5), Mudde (2004, 2017) offers a different explanation that it is not a full worldview or discourse but a “thin-centered” ideology separating society into two groups, which “considers society to be ultimately separated into two homogeneous and antagonistic camps, ‘the pure people’ versus ‘the corrupt elite’ which argues that politics should be an expression of the… general will of the people” (p. 543). Opposed to this theory’s “thin-centered” ideology, “thick-centered” ideologies include concepts such as fascism, liberalism, socialism, etc., which can stand on their own. “Thin-centered” populism lacks its own tenants and ideological points. It attaches to or is assimilated into other ideologies like a parasite to promote them to the public but is unable to stand alone (Mudde & Kaltwasser, 2017, p. 6; Moffitt, 2020). Mudde and Kaltwasser (2017) admit this as a potential limitation of the theory; while flexible and applicable to many political styles and ideologies, it is broad and therefore only useful when it includes the binary of common people/corrupt elites and a thin-centered ideology to attach to another, excluding everything else. As such, one cannot be a “populist” without another espousing a distinct ideology.

There are multiple advantages to this theory as well as reasons it is applicable to this study. These include that the theory describes clear boundaries between populism and “non-populism,” allows for the construction of categories of populism in relation to other concepts and forms of governance, is able to cross national or regional contexts, and is versatile, as it is able to be used at different levels of analysis (Mudde, 2017). Thus, this theory states that populism can be flexible and take various shapes and forms in different times, places, and contexts as it combines with different ideologies and systems.

Whatever form it takes, however, Mudde and Kaltwasser (2017) argue populism has a dark side; the element of a general will of the people can legitimize authoritarianism and justify attacks on outgroup members. In Mudde and Kaltwasser’s (2017) view, this threatens the homogeneity of the people rather than, as in the popular agency approach Laclau (2005) describes, acting as a positive, democratizing force. These elements manifest in the creation of perceptions of others. Populist communication often relies on encouraging followers to blame unwanted circumstances or outcomes on another’s actions, framing in- and out-groups based on norms and stereotypes of their own and framing competing groups as scapegoats to blame or punish, often exploiting the emotions that can influence decision making, such as fear, anger, or patriotism (Dimitrov et al., 2021; Hawkins et al., 2018; Marlin, 2002; Rahn, 2019).

2.2. Propaganda

Populist communicators rely on maintaining their ideological narratives, which inevitably become part of their lasting propaganda (Waisbord, 2018). Just as the literature above poses the question of whether populism is good or bad for democracy and answers it with “it depends,” this applies to propaganda, too. Also similar to populism’s position in past scholarship, the elusive definition of propaganda has been debated and often overlaps (Guess & Lyons, 2020), though its various individual techniques, attributes, and types have been studied at length.

At the beginning of his book, aptly entitled Propaganda, Bernays (1928) stated, “We are governed, our minds are molded, our tastes formed, our ideas suggested, largely by men we have never heard of. This is a logical result of the way in which our democratic society is organized” (p. 9). This ties into both the concept of the elite as the enemy of populism as well as ways populist leaders mobilize their followers.

Henderson (1943) explains the challenge of defining propaganda due to the emotional, subjective nature of the word. This is similar to Bernays’s (1928) claim there is “no word in the English language… whose meaning has been so sadly distorted as the word ‘propaganda’” (p. 21). He continues, stating it “becomes vicious and reprehensive only when its authors consciously and deliberately disseminate what they know to be lies, or when they aim at effects which they know to be prejudicial to the common good” (Bernays, 1928, p. 22). Walton (1997) states the deceptiveness of propaganda is due to the format within which most propaganda is presented.

Herman and Chomsky (1988) introduced a macro-level propaganda model that, alongside work by other scholars (e.g., Bernays, 1928; Lee & Lee, 1939), acts as a starting point in studying propaganda and represents an important paradigm of propaganda studies in the field of mass communication. Their model focuses on inequalities of power and wealth and their systemic influence on mass media, government, and private actors. They claim this influence is in how mass media, government, and private actors select their messages, set the agenda of discourse, and even dominate the information environment (Herman & Chomsky, 1988). This focus on power dynamics is a key point in understanding populist communicators’ use of strategic propaganda narratives. While Herman and Chomsky’s model is influential, the present study adopts an approach at more of a micro-scale than Herman and Chomsky’s macro one, focusing mainly on the message-level of analysis to understand propaganda.

These perspectives, in combination with Benkler et al.’s (2018) definition featured in the introduction, demonstrate how the usage of political propaganda is a deliberate attempt to alter, maintain, and otherwise control a balance of political power advantageous to the propagandist (Jowett & O’Donnell, 2015). Further, they show it is a form of politically ideological communication (Hyzen, 2021; Jang, 2013) that makes deliberate use of persuasive appeals or techniques, as well as a systematic dissemination of one-sided, biased information (Walton, 1997; O’Shaughnessy, 1996), to exploit emotions or emotional fallacies (Postman, 1979; O’Shaughnessy, 1996; Jowett & O’Donnell, 2015) and manipulate rationality. This is for the purpose of managing collective attitudes and shaping perceptions, cognition, and behavior (Fawkes, 2007).

Therefore, the goal of propaganda is often understood as the manipulation of behavior and attitudes to instill values that advance a cause linked to an institutional ideology and objective (Jowett & O’Donnell, 2015; Denisenko, 2021; Hyzen, 2021). Further, propaganda is used in order to influence opinions in a target audience and promote an ideological cause, point of view, or goal to the benefit of the originator or detriment of the target via mass or direct media channels. By this understanding, it can fit as another companion component in populism’s thin-centered ideology in how it hangs onto other belief or governing systems. In fact, many populist leaders, such as Viktor Orban of Hungary or Aleksandr Lukashenko of Belarus, have been able to capture and control media outlets for the dissemination of their populist propaganda (Ribeiro, 2025). Donald Trump’s actions can be interpreted as attempting to do the same, using his position to influence any source he can to spread his form of propaganda.

In Lee and Lee’s research (1939) for the Institute of Propaganda Analysis, seven propaganda devices or “tricks of the trade” were introduced that allow for a more detailed analysis of propaganda (Marlin, 2002; Sproule, 2001; Hobbs & McGee, 2014; Denisenko, 2021; Clark, 1997; Maliukevičius, 2008). These include (1) name calling, (2) glittering generalities, (3) transfer, (4) testimonial, (5) plain folks, (6) card stacking, and (7) bandwagoning (Sproule, 2001, p. 136).

To examine the validity of this argument, this study seeks to explore if and how these tactics are used in Trump’s messaging or if any of them are outdated or less irrelevant. As such, through a deductive content analysis, the present study seeks to identify in what ways, if any, Donald Trump’s messaging utilizes these tactics, and which groups or individuals manifest as targets.

2.2.1. The Classification of Propaganda Techniques

The following sections serve two purposes. First, they describe each tactic in detail. Second, they outline the inclusion criteria for each category when coding the data.

Name Calling: “Lyin’ Kamala and Sleepy Joe!”

One propagandistic technique is name calling, including labeling the target as “something the target audience fears, hates, finds undesirable, or loves” (Dimitrov et al., 2021, p. 6606). Additionally, name calling includes ad hominem attacks (Cunningham, 2002) or ridiculing individuals, either with disparaging nicknames or mocking their general actions in ways often unsupported by rational argument or evidence (Maliukevičius, 2008). Further, this technique includes the use of sarcasm and negative symbols against opponents (Maliukevičius, 2008). Creating these stigmatizing names and “unattractive labels” promotes hate, anger, and fear for “individuals, groups, nations, races, policies, practices, beliefs, and ideas which [the originator] would have us condemn and reject” (Sproule, 2001, p. 136). This is done to discredit the target or make the target audience accept a conclusion without full cognition or consideration of essential facts (Hobbs & McGee, 2014). This method does not necessarily aim to create support for the propagandist but rather lower support for the target, similar to how criticizing one party does not increase support or improve views for the other party (Fisher, 2020).

Glittering Generalities: “Make America Great Again!”

Glittering generalities are associations between the propaganda and “virtue words” such as “truth, freedom, honor, liberty,” etc., in order to make people form a thoughtless judgment under the influence of an emotional impression (Sproule, 2001). A populist would utilize this technique by calling for a return to a nostalgic history before the present was ruined by elites. This technique involves an attempt to “sway emotions through the use of shining ideals or virtues, such as freedom, justice, truth, education, [and] democracy in a large, general way” (Hobbs & McGee, 2014, p. 59). These ideas give power to the populist leader or propagandist exploiting them (Marlin, 2002) and to the propaganda surrounding individuals, their actions, or general ideas despite being unsupported by evidence or rational arguments (Maliukevičius, 2008; Marlin, 2002). Denisenko (2021) explains relevant myths commonly used by propagandists, including the myth of a “golden age” (“Make America Great Again!”), the savior myth, and the unity myth. The savior myth is often used in political or war propaganda, connected to the image of a heroic figure who comes to help in a hard time, while the unity myth emphasizes a fear of division or difference of opinion, seeking a common belief, and striving for great collective aspirations (Giradet, 2007, as cited in Denisenko, 2021, pp. 35–36). An example of this is the “manifest destiny” of the United States.

Transfer

The original definition reads: “Transfer carries the authority, sanction, and prestige of something respected and revered over to something else in order to make the latter acceptable” (Marlin, 2002, p. 103). This technique involves the comparison of positive or negative phenomena with the object of propaganda (Maliukevičius, 2008). This includes a method called “flag waving,” wherein the propagandist uses a symbol that the target recognizes and respects (Hobbs & McGee, 2014), such as a flag, making use of the attributes a group perceives it to espouse. Populist propagandists are able to transfer their authority or prestige to something they would have their followers respect, such as transferring reverence for national or religious symbolism to a political project (Sproule, 2001).

Testimonials: “He Will Not Let You Down!”

The testimonials technique utilizes authoritative opinion or expert recommendations to reinforce the impression, idea, or ideology promoted and desired by the propagandist (Maliukevičius, 2008). It includes “getting not only good, plain, solid citizens, but also social and business leaders to endorse the party or candidate” (Hobbs & McGee, 2014, p. 59). Testimonials link an idea or program to a specific, favored person or institution (Sproule, 2001). These expert testimonials, recommendations, and authoritative opinions are used to reinforce the intended impression of the propaganda material itself, as well as its source (Maliukevičius, 2008). It can also take the form of the propagandist offering their own testimonial or endorsement to another.

Plain Folks: “I’m Just Like You!”

This technique is inherently tied to the key definition of populism in that it demonstrates how the propagandist is (or claims to be) just like the rest of the common people, or “plain folk”: a simple representative of the entire people, in the name of the people, against an elite (Hobbs & McGee, 2014; Maliukevičius, 2008). Populists can present themselves or their representatives as “plain folks” or appeal to the “common man” to establish a group identity with the people. This is done in order to create a favorable impression about their ideas, portray themselves as a representative of the nation, and to present the propaganda material or event as the aspiration of the rest of the people; you can trust what they say is “good” because they are, in a way, like a friend (Sproule, 2001; Marlin, 2002).

Card Stacking: Political Smokescreens and Historical Revisionism

The original definition of the sixth tactic, card stacking, reads as, “The selection and use of facts and falsehoods, illustrations or distractions, and logical or illogical statements in order to give the best or the worst possible case for an idea, program, person, or product” (Marlin, 2002, p. 105). Card stacking manifests when the propagandist metaphorically “stacks the cards” against the actual facts (Hobbs & McGee, 2014) by presenting positive information about a person or phenomenon while withholding negative information, or vice versa, to manipulate a narrative (Maliukevičius, 2008, p. 84). Card stacking over- or under-emphasizes certain ideas to distort information, put a beneficial spin on it, or omit it entirely, acting as a “smokescreen” that causes audiences to forget or not see inconvenient information and accept half-truths and presenting positive information but withholding negative information or vice versa (Sproule, 2001; Maliukevičius, 2008). An example of this is manipulating or revising history by placing created or half-truths against facts (Chakars & Ekmanis, 2022).

Bandwagon: “Everyone’s Doing It!”

Finally, the seventh tactic is the bandwagon. This technique of political peer pressure is used to seize emotions and make people both follow the “political Pied Pipers” and bring others along with them (Hobbs & McGee, 2014, p. 59). Hobbs and McGee (2014) describe this simply as “everybody’s doing it,” as it influences people to “follow the crowd,” appealing to the desire to behave like everyone else and follow the current fashion (Sproule, 2001; Maliukevičius, 2008).

2.2.2. Populist Propaganda on Social Media

As social media has become, according to Dimitrov et al. (2021), the main communication channel and primary source of news for many people, it has also become a tool for political actors to spread computational propaganda in well-determined communities in order to influence opinions and behavior. This can be done in textual social media posts on X/Twitter, Facebook, Reddit, Truth Social, or elsewhere, as well as in visual formats such as internet memes. The sheer number of these available platforms makes it easy for misinformation, propaganda, and disinformation to go viral and spread to wide groups quickly (Dimitrov et al., 2021).

Twitter was Trump’s main communication platform during his first presidential election campaign, where he used a distinct “direct and unapologetic form of self-promotion” that featured both promotion and criticism (Clarke & Grieve, 2019). Trump uses online platforms to “express opinions, attack opponents, and promote [his] campaign” rather than “debating his critics or disputing their claims… doubling down on controversial views” (p. 20).

As Trump’s online messaging explicitly distinguishes between in- and outgroup members (Luo et al., 2021), Mudde and Kaltwasser’s (2017) approach to populism, which focuses on the division between the plain folks and corrupt Washington elites, becomes particularly salient. This framework helps contextualize the present study’s coding scheme of propaganda techniques within perceptions to both political allies and opponents in populist political communication online.

It is relevant to this study to note social media allows the originator of propaganda to tailor content to specific targets (Guess & Lyons, 2020). Indeed, as Siegel (2020) explains, harmful rhetoric and hate speech inciting harm and extremism has become increasingly more prevalent on mainstream social media platforms. Online propaganda, when shared within groups, can create echo chambers that reinforce extreme political opinions and increase political polarization (Barbera, 2020). Such polarization can then have offline consequences such as ethnic violence (Siegel, 2020).

Guess and Lyons (2020) argue that in order to understand such online behavior and targeting online, it is important to move beyond the macro-level analysis of media ecosystems. Thus, the present study takes a micro approach, examining Donald Trump’s online rhetoric in a descriptive way in order to analyze his usage of classic propaganda tactics.

2.2.3. Contemporary Applicability

The Institute for Propaganda Analysis (IPA), operating from 1937–1942, and its propaganda tactics emerged during a particular historical context that reflected normative assumptions of American democracy, attempting to distinguish between “good” and “bad” propaganda (Bauer, 2024). Bauer (2024) argues that while researchers of propaganda must “acknowledge and defend the political visions and values that drive [their] work,” they must also recognize that such frameworks are not inherently neutral and should be used with awareness of historical context.

With this historical framework in mind, and although this framework is nearly a century old, the seven tactics are still effective descriptors of propagandistic actions. According to Maliukevičius (2008), novel propaganda techniques are often derivatives or variations of these traditional tactics, allowing them to stay relevant as a foundation of propaganda studies, even in the twenty-first century (p. 85). Hobbs and McGee (2014) echo this assertion, stating that these tactics have endured as a “dominant approach to explore persuasion and propaganda,” including on the modern internet.

Although the techniques that influence public opinion originate from classical propaganda studies, Papathanassopoulos and Giannouli (2025) describe a new era for political communication marked by increased mediatization set mainly in social media. Further, the more modern digital political communication scholarship explains that public discourse, including during political campaigns and election periods more specifically, is not only influenced by message content but specific platform affordances that allow actions that were difficult to achieve using older mass media technology (Theocharis et al., 2023; Bossetta, 2024). Social media platforms are now integrated into all elements of political participation as a decentralized, personalized, and more interactive tool or stage allowing the quick and wide sharing of information, as well as an arena of discussion and debate (Theocharis et al., 2023; Papathanassopoulos & Giannouli, 2025). Bossetta (2024) states that the complexity of online participation and political campaigning is due to the varied nature of social media platforms and the utility they provide. Due to this fact, “each platform usually serves a particular function or role within that strategy” (Bossetta, 2024, p. 2). These different uses affect how political communication takes place on social media, ultimately deciding what content gains visibility, how it is produced and shared, or if it is even allowed to be posted at all (Bossetta, 2024). Because of this, the use of the classical propaganda techniques may present differently across separate platforms and the context of their varying digital communication environments during political campaigns. Papathanassopoulos and Giannouli (2025) state that this has “motivated politicians to adopt communication strategies that align with the logic of social media: highly personalized, emotionally charged, and often conflict-driven narratives… frequently at the expense of more traditional, policy focused discourse” (pp. 4–5).

Seo (2020) states that “comparing propaganda tactics used in traditional media with those used in social media can provide useful context” in understanding online propaganda in the digital age, a contribution of the present study. Building on what Maliukevičius (2008), Hobbs and McGee (2014), and Seo (2020) say, the elements of the IPA propaganda tactics remain relevant in the internet age and are found within Donald Trump’s online rhetoric in his tailored messaging, as evidenced in the present study’s data and findings. Thus, the classical propaganda techniques of the IPA remain relevant amidst the new affordances of the evolving digital landscape, and these categories become a useful framework for analyzing modern digital propaganda and serve as the codes for this analysis, to be expounded upon in the Section 3 below.

2.3. Research Questions and Hypotheses

With the literature and scholarship in mind, this study poses the following research questions and corresponding hypotheses:

• RQ1: In what ways, if any, has Donald Trump utilized each of the seven classic propaganda techniques, and if used, what groups did he use them against in posts on Truth Social during his 2024 American presidential election campaign?

H1a. 

Donald Trump’s messaging has a positive relationship with name calling when targeting outgroups.

H1b. 

Donald Trump’s messaging has a positive relationship with positive techniques (glittering generalities, bandwagon, testimonials) when targeting ingroup members and a negative relationship with outgroups.

H1c. 

As flexible propaganda techniques that can be used with in- or out-group members, card stacking and transfer are associated with both groups but have weak target-specific relationships.

• RQ2: In what ways, if any, were emotional appeals used in Donald Trump’s posts on Truth Social during his 2024 American presidential election campaign?

H2a. 

Donald Trump’s messaging has a negative relationship with valence.

H2b. 

Donald Trump’s messaging has a positive relationship with using emotional appeals.

For an explanation of H1a–c, name calling posts are expected to be overwhelmingly negative, but glittering generalities, bandwagon, and testimonials are generally used in subjectively positive ways, while card stacking and transfer are more flexible and able to be used in both positive and negative ways.

3. Materials and Methods

3.1. Jowett and O’Donnell’s Plan of Propaganda Analysis

Cunningham (2002) states that some writers insist “the definition of propaganda must be method-driven as opposed to the practice of assembling descriptions and positing mere ‘content definitions’” (p. 80). Therefore, the terms outlined above will serve as the variables in this deductive content analysis. As a content analysis, the variables were coded based on the presence of each tactic as described by the literature as a rhetorical device rather than on specific linguistic markers.

Likewise, Jowett and O’Donnell (2015) offer a ten-step plan of analysis for a framework to study propaganda. It helps in identifying the key components of the communication and lays out the study effectively.

The first step in this content analysis, according to Jowett and O’Donnell (2015), is to “identify the ideology and purpose of the propaganda campaign” (p. 314). This ideology is the thin-centered populism that manifests in the seven propaganda tactics found within Donald Trump’s messaging. The second step is to “describe the context in which the propaganda occurs” (p. 314). The context of this study occurs during Trump’s 2024 American presidential election campaign and includes every textual post he made during that time, excluding images and videos. Step three is to identify the propagandist. In this case, Donald Trump serves as the sole propagandist, as it is his Truth Social posts that serve as the corpus of data in this analysis. Step four is to describe the structure of the propaganda organization. While not the focus of this study, the organization is the Republican party and powers-that-be supporting Trump’s campaign, which leads into step five, identifying the target audience, which includes the MAGA crowd and other supporters. Step six is a description of “media utilization techniques” (p. 314). In this case, Donald Trump made over 1300 posts on his social media platform, Truth Social. While many forms of propaganda are visual, Cunningham (2002) states that much of it is linguistically formatted, such as in social media posts. Connected to this is step seven, to describe the “special techniques to maximize effect” (Jowett & O’Donnell, p. 314). The special techniques used here will be measured through the seven propaganda techniques described above. Step eight describes analyzing audience reaction to various techniques, while step nine describes any present counterpropaganda. Step ten is to evaluate the effects of propaganda. Audience reaction such as attitudinal or behavioral change or counter-propaganda responses as well as other persuasive effects are not analyzed here. Therefore, these three steps are outside the scope of the current study and will not be included. However, as this study focuses on the tactics and targets rather than the causal effects of messaging, this does not limit the findings of this study.

3.2. Variables

3.2.1. Dependent Variables

The dependent variables for this study include the various targets mentioned in Donald Trump’s Truth Social posts. An inductive approach was used for these variables to uncover the various categorizations of targets. The following categories were built to classify targets: Democrats, Republicans, the media, institutions, immigrants, and foreign leaders. The “Democrats” outgroup includes Democratic leaders, politicians, supporters, and progressive individuals across the center-to-left side of the political spectrum. Similarly, “Republicans” include Republican leaders, politicians, supporters, and other conservatives across the center-to-right side of the political spectrum. However, this also includes both ingroup (supporters) and outgroup members perceived by Trump as traitors or “RINOs” (Republicans in name only). “The Media” includes both pro-Trump, conservative media and more liberal leaning, opposition media. “Institutions” includes corporations, businesses, banks, universities, NGOs, and nonprofit organizations, but not media outlets. “Immigrants” includes both documented and undocumented immigrants.

These are used as dependent rather than independent variables, as my analysis is centered around the question of “If X tactic is used, what Y group is most likely to be targeted?”

3.2.2. Independent Variables

The independent variables for this study include the seven propaganda techniques. Posts were analyzed with a deductive approach wherein pre-existing categories were used to code. These, as explained and operationalized above, include name calling, glittering generalities, plain folks, transfer, testimonials, bandwagon, and card stacking.

I also include several other independent variables and covariates. These include the presence of propaganda (generally), the presence of emotional appeals (generally), and the related variables of anger appeals, fear appeals, and appeals to positive emotions such as patriotism, nostalgia, joy, etc.

While intercoder reliability is a best practice in content analysis, the size of the dataset (n = 1319) and overall number of variables (23, although some were not ultimately included in the study) resulted in a substantial number of coding decisions that required familiarity with wider propaganda research and context. Therefore, as a result, a codebook was created and analysis performed by a single coder, which may limit generalizability. This is a limitation of the present study, and similar research in the future would benefit from utilizing multiple coders.

3.3. Inclusion and Exclusion Criteria

Longer explanations of each code were given by the authors cited above (e.g., Dimitrov et al., 2021; Hobbs & McGee, 2014; Lee & Lee, 1939; Maliukevičius, 2008, etc.), who provided inspiration for the coding criteria outlined in

Table 1. Inclusion and exclusion criteria (independent variables).

Variable	Coding Criteria
Name calling	Posts that used nicknames or insults that smear a reputation or intend to create doubt in someone’s expertise.
Glittering Generalities	Posts that contained the use of positive slogans, words, or symbols attached to campaign promises, issues, people, places, groups, or objects.
Transfer	Posts that make positive or negative associations between two objects. The first object is something the poster and audience may admire or despise, and the second object is the entity being targeted.
Testimonial	Posts praising someone or accepting praise from someone else.
Plain folks	Posts that contain language about fighting the elite on behalf of the people. Includes any posts depicting the poster as one of the common people.
Card stacking	Posts that downplay positive/negative subjects while focusing on the opposite.
Bandwagon	Posts with calls to join a party, rally, or event because everyone else is. Language may include generalized statements such as “everyone is doing it” or “everyone is.”
Emotional appeals	Any post appealing to fear, anger, disgust, pride, joy, patriotism, etc. Combines appeals to anger, fear, and positivity categories.
Appeals to anger	Posts with negative valence and language intended to increase anger, such as “you should be mad!” or “get angry!” as well as describing events that directly and negatively impact the audience, such as “they’re stealing your jobs!”
Appeals to fear	Posts with negative valence and language, creating support for something by creating fear or anxiety about an alternative.
Positivity appeals	Posts with positive valence, with language appealing to joy, pride, or patriotism.
Positive valence	Posts with a positive, encouraging, and uplifting tone, emphasizing good things.
Negative valence	Posts with a negative, rude, or disparaging tone, emphasizing bad things.
Neutral valence	Posts without positive or negative appeals. Statements.

. Many posts included multiple tactics or targets, which did not interfere with analysis.

3.4. Models

Logit (or logistic regression) models were used as the estimator to find the relationship between binary independent and dependent variables as well as to provide the predicted probabilities, odds ratios, and statistical significance of each relationship. Descriptions of independent and dependent variables are provided below. This type of test is appropriate because of the binary coding of each variable (0 = not present, 1 = present). Only one variable, valence, was coded differently (0 = negative, 1 = positive, 2 = neutral). Valence was only coded for direction, not arousal or intensity.

Logit Model Estimators (Equations (1)–(11))

RQ1: The logit estimators’ equations for RQ1’s hypotheses are as follows, and Y = the various targets described below (Democrats, Republicans, media, institutions, immigrants, or foreign leaders):

H1a,b,c. 

Equations (1)–(6): Where Y = Democrat, Republican, media, institutions, immigrants, or foreign leaders, respectively.

Equation (1): Democrats as the target

$$\begin{array}{l}\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{P}\mathrm{r}\left(Y_{Democrats,i}=1\right)}{1-\mathrm{P}\mathrm{r}\left(Y_{Democrats,i}=1\right)}}\right)\\ ={\beta }_0+{\beta }_1\left({\mathrm{N}\mathrm{a}\mathrm{m}\mathrm{e}\mathrm{C}\mathrm{a}\mathrm{l}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)+{\beta }_2\left({\mathrm{G}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{e}\mathrm{s}}_i\right)+{\beta }_3\left({\mathrm{T}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{f}\mathrm{e}\mathrm{r}}_i\right)+{\beta }_4\left({\mathrm{T}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{s}}_i\right)+{\beta }_5\left({\mathrm{P}\mathrm{l}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{l}\mathrm{k}\mathrm{s}}_i\right)\\ + {\beta }_6\left({\mathrm{B}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{w}\mathrm{a}\mathrm{g}\mathrm{o}\mathrm{n}}_i\right)+{\beta }_7\left({\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)\end{array}$$

(1)

Equation (2): Republicans as the target

$$\begin{array}{l}\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{P}\mathrm{r}\left(Y_{Republicans,i}=1\right)}{1-\mathrm{P}\mathrm{r}\left(Y_{Republicans,i}=1\right)}}\right)\\ ={\beta }_0+{\beta }_1\left({\mathrm{N}\mathrm{a}\mathrm{m}\mathrm{e}\mathrm{C}\mathrm{a}\mathrm{l}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)+{\beta }_2\left({\mathrm{G}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{e}\mathrm{s}}_i\right)+{\beta }_3\left({\mathrm{T}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{f}\mathrm{e}\mathrm{r}}_i\right)+{\beta }_4\left({\mathrm{T}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{s}}_i\right)\\ + {\beta }_5\left({\mathrm{P}\mathrm{l}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{l}\mathrm{k}\mathrm{s}}_i\right)+{\beta }_6\left({\mathrm{B}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{w}\mathrm{a}\mathrm{g}\mathrm{o}\mathrm{n}}_i\right)+{\beta }_7\left({\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)\end{array}$$

(2)

Equation (3): The media as the target

$$\begin{array}{l}\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{P}\mathrm{r}\left(Y_{Media,i}=1\right)}{1-\mathrm{P}\mathrm{r}\left(Y_{Media,i}=1\right)}}\right)\\ ={\beta }_0+{\beta }_1\left({\mathrm{N}\mathrm{a}\mathrm{m}\mathrm{e}\mathrm{C}\mathrm{a}\mathrm{l}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)+{\beta }_2\left({\mathrm{G}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{e}\mathrm{s}}_i\right)+{\beta }_3\left({\mathrm{T}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{f}\mathrm{e}\mathrm{r}}_i\right)+{\beta }_4\left({\mathrm{T}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{s}}_i\right)\\ + {\beta }_5\left({\mathrm{P}\mathrm{l}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{l}\mathrm{k}\mathrm{s}}_i\right)+{\beta }_6\left({\mathrm{B}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{w}\mathrm{a}\mathrm{g}\mathrm{o}\mathrm{n}}_i\right)+{\beta }_7\left({\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)\end{array}$$

(3)

Equation (4): Institutions as the target

$$\begin{array}{l}\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{P}\mathrm{r}\left(Y_{Institutions,i}=1\right)}{1-\mathrm{P}\mathrm{r}\left(Y_{Institutions,i}=1\right)}}\right)\\ ={\beta }_0+{\beta }_1\left({\mathrm{N}\mathrm{a}\mathrm{m}\mathrm{e}\mathrm{C}\mathrm{a}\mathrm{l}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)+{\beta }_2\left({\mathrm{G}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{e}\mathrm{s}}_i\right)+{\beta }_3\left({\mathrm{T}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{f}\mathrm{e}\mathrm{r}}_i\right)+{\beta }_4\left({\mathrm{T}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{s}}_i\right)\\ + {\beta }_5\left({\mathrm{P}\mathrm{l}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{l}\mathrm{k}\mathrm{s}}_i\right)+{\beta }_6\left({\mathrm{B}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{w}\mathrm{a}\mathrm{g}\mathrm{o}\mathrm{n}}_i\right)+{\beta }_7\left({\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)\end{array}$$

(4)

Equation (5): Immigrants as the target

$$\begin{array}{l}\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{P}\mathrm{r}\left(Y_{Immigrants,i}=1\right)}{1-\mathrm{P}\mathrm{r}\left(Y_{Immigrants,i}=1\right)}}\right)\\ ={\beta }_0+{\beta }_1\left({\mathrm{N}\mathrm{a}\mathrm{m}\mathrm{e}\mathrm{C}\mathrm{a}\mathrm{l}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)+{\beta }_2\left({\mathrm{G}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{e}\mathrm{s}}_i\right)+{\beta }_3\left({\mathrm{T}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{f}\mathrm{e}\mathrm{r}}_i\right)+{\beta }_4\left({\mathrm{T}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{s}}_i\right)\\ + {\beta }_5\left({\mathrm{P}\mathrm{l}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{l}\mathrm{k}\mathrm{s}}_i\right)+{\beta }_6\left({\mathrm{B}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{w}\mathrm{a}\mathrm{g}\mathrm{o}\mathrm{n}}_i\right)+{\beta }_7\left({\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)\end{array}$$

(5)

Equation (6): Foreign leaders as the target

$$\begin{array}{l}\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{P}\mathrm{r}\left(Y_{ForeignLeaders,i}=1\right)}{1-\mathrm{P}\mathrm{r}\left(Y_{ForeignLeadersi}=1\right)}}\right)\\ ={\beta }_0+{\beta }_1\left({\mathrm{N}\mathrm{a}\mathrm{m}\mathrm{e}\mathrm{C}\mathrm{a}\mathrm{l}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)+{\beta }_2\left({\mathrm{G}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{e}\mathrm{s}}_i\right)+{\beta }_3\left({\mathrm{T}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{f}\mathrm{e}\mathrm{r}}_i\right)+{\beta }_4\left({\mathrm{T}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{s}}_i\right)\\ + {\beta }_5\left({\mathrm{P}\mathrm{l}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{l}\mathrm{k}\mathrm{s}}_i\right)+{\beta }_6\left({\mathrm{B}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{w}\mathrm{a}\mathrm{g}\mathrm{o}\mathrm{n}}_i\right)+{\beta }_7\left({\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{d}\mathrm{S}\mathrm{t}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}}_i\right)\end{array}$$

(6)

RQ2: The logit models’ equations for RQ2’s hypotheses are as follows, with Y = the presence of an emotional appeal in Equation (1), fear appeal in Equation (7), anger appeal in Equation (8), positivity appeal in Equation (9), and valence in Equation (10).

H2a. 

Equations (7)–(10): Y_i = emotional appeal, fear appeal, anger appeal, or positivity appeal, respectively. (Y_i = 0, 1 for presence of each type of appeal).

$$\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{Pr}(Y_{Emotion,i}=1)}{1-\mathrm{Pr}(Y_{Emotion,i}=1)}}\right)={\beta }_0+{\beta }_1(\mathrm{P}\mathrm{r}\mathrm{o}\mathrm{p}\mathrm{a}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{d}{\mathrm{a}}_i)$$

(7)

$$\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{Pr}(Y_{Fear,i}=1)}{1-\mathrm{Pr}(Y_{Fear,i}=1)}}\right)={\beta }_0+{\beta }_1(\mathrm{P}\mathrm{r}\mathrm{o}\mathrm{p}\mathrm{a}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{d}{\mathrm{a}}_i)$$

(8)

$$\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{Pr}(Y_{Anger,i}=1)}{1-\mathrm{Pr}(Y_{Anger,i}=1)}}\right)={\beta }_0+{\beta }_1(\mathrm{P}\mathrm{r}\mathrm{o}\mathrm{p}\mathrm{a}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{d}{\mathrm{a}}_i)$$

(9)

$$\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{Pr}(Y_{Positivity,i}=1)}{1-\mathrm{Pr}(Y_{Positivity,i}=1)}}\right)={\beta }_0+{\beta }_1(\mathrm{P}\mathrm{r}\mathrm{o}\mathrm{p}\mathrm{a}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{d}{\mathrm{a}}_i)$$

(10)

H2b. 

Equation (11): Where Y_i = valence (coded y = 0, 1, 2 for negative, positive, neutral) as a multinomial regression generalized linear model, as it was the only variable not coded as a binary. k = positive or neutral valence. Propaganda_i refers to if a propaganda tactic was present to predict the valence of propaganda messaging.

$$\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{\mathrm{Pr}(Y_{Valence,i}=k)}{\mathrm{Pr}(Y_{Valence,i}=0)}}\right)={\beta }_{0k}+{\beta }_{1k}(\mathrm{P}\mathrm{r}\mathrm{o}\mathrm{p}\mathrm{a}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{d}{\mathrm{a}}_i)$$

(11)

3.5. LLM Statement

ChatGPT 5.5 was used to assist with R code and to model equations but was not used for any language within the paper.

4. Results

4.1. Targets and Tactics

The second research question had two parts: (1) what groups were mentioned in Trump’s posts, and (2) what propaganda tactics did he use against them? Table 1 shows the raw frequencies and proportions corresponding to both group and tactic in each of Trump’s posts (n = 1319). Targets fell into six main categories: Democrats, Republicans (both ingroup and outgroup Republicans), the media, institutions (such as NGOs, international organizations, banks, companies, etc.), immigrants, and foreign leaders. Every tactic was used to some extent against every group. (See

Table A1. Logistic regression predicting mentions of Democrats

Predictor	β	SE	Wald χ2	df	p	OR
(Intercept)	−1.892	0.158	143.39	1	<0.001	0.151
Name Calling	2.721	0.157	300.30	1	<0.001	15.190
Glittering Generalities	−0.503	0.167	9.10	1	0.003	0.605
Transfer	−0.076	0.219	0.12	1	0.728	0.927
Testimonials	−0.583	0.208	7.86	1	0.005	0.558
Plain Folk	0.494	0.251	3.88	1	0.049	1.640
Bandwagon	0.328	0.329	0.99	1	0.320	1.388
Card Stacking	1.468	0.162	81.85	1	<0.001	4.338

in Appendix A for all frequencies.)

For the following analysis, logit models as explained above were used to examine how the propaganda tactics were used to predict the target of Trump’s social media posts.

4.1.1. Democrats

First, Democrats were by far the most targeted group, as seen in

Table 2. Frequencies of target and tactic.

Truth Social Posts N = 1319	Democrats N = 648 (39.12%)	Republicans N = 370 (28.05%)	Media N = 186 (14.1%)	Institutions N = 225 (17.06%)	Immigrants N = 232 (17.6%)	Foreign Actors N = 65 (4.92%)
Name Calling N = 618 (46.85%)	516 (79.63%)	130 (35.14%)	120 (64.52%)	76 (33.8%)	108 (46.55%)	38 (58.46%)
Glittering Generalities N = 579 (43.9%)	235 (36.27%)	247 (66.76%)	61 (33%)	160 (71.11%)	162 (69.83%)	35 (53.85%)
Transfer N = 260 (19.71%)	151 (23.3%)	130 (35.14%)	24 (12.9%)	99 (44%)	126 (54.31%)	21 (32.31%)
Testimonials N = 285 (21.61%)	68 (10.5%)	175 (47.3%)	46 (24.73%)	125 (55.6%)	105 (45.26%)	7 (10.77%)
Common Folk N = 155 (11.75%)	97 (15%)	43 (11.5%)	14 (7.53%)	23 (10.22%)	39 (16.81%)	10 (15.38%)
Bandwagon N = 85 (6.44%)	44 (6.8%)	29 (7.84%)	13 (7%)	14 (6.22%)	10 (4.31%)	5 (7.7%)
Card Stacking N = 765 (58%)	507 (78.24%)	245 (66.22%)	116 (62.4%)	170 (75.6%)	189 (81.47%)	53 (81.54%)

. The two strongest positive predictors of Trump’s posts targeting a Democrat are name calling and card stacking. The data show that when Trump engages in name calling or ad hoc attacks, it is approximately 15 times more likely that he is targeting a Democrat than other groups (OR = 15.19, p < 0.001). Likewise, card stacking as a tactic predicts Democrats as the target to be approximately four times more likely (OR = 4.34, p < 0.001). Another less significant result is that when appealing to the plain folk, he is 1.6 times more likely to be targeting Democrats (OR = 1.64, p = 0.05).

On the other hand, the two strongest negative predictors include testimonials (OR = 0.56, p = 0.005) and glittering generalities (OR = 0.61, p = 0.003). This means that he is about 44% less likely to use a testimonial and 39% less likely to use glittering generalities when Democrats are the subject of a post. Using transfers (OR = 0.93, p = 0.73) and calls to “join the bandwagon” (OR = 1.39, p = 0.32) were not found to be significant and thus mean they were used against Democrats but not disproportionately so.

4.1.2. Republicans

The three strongest positive predictors of Trump’s posts targeting or featuring a Republican are glittering generalities, testimonials, and card stacking. First, the use of testimonials was the strongest predictor, meaning that if Trump is offering an endorsement or testimonial, it is 5.4 times more likely to be going to a Republican ally. Second, when using glittering generalities, Trump is 2.5 times more likely to be using them to praise political allies (OR = 2.56, p < 0.001). Finally, the third strong predictor that he is mentioning Republicans is card stacking (OR = 2.05, p < 0.001). Other less statistically significant results include transfer (OR = 1.81, p = 0.001) and name calling, where he is less likely to be targeting his own party (OR = 0.64, p = 0.006). However, he still does target those in his party he perceives to be traitors to the party, or “RINOs,” with some frequency. Appeals to the plain folk had a negative association, and bandwagon had a positive one, but neither were found to be statistically significant, meaning they are not disproportionately used against Republicans.

4.1.3. The “Fake News” Media

The two strongest positive predictors of Trump targeting the media are name calling and testimonials. Name calling made the likelihood of targeting the media 2.77 times more likely (OR = 2.77, p < 0.001), and testimonials indicated the media would be 2.18 times more likely to be the target.

The two negative predictors were glittering generalities and transfer. The data show Trump is approximately 30% less likely to use glittering generalities (OR = 0.7, p = 0.05) and 51% less likely to use transfer techniques (OR = 0.49, p = 0.005) against the media. The tactics of plain folk, bandwagon, and card stacking were not found to be significant predictors of media as the target.

4.1.4. Institutions

The four strongest positive predictors of Trump’s posts targeting institutions include glittering generalities, transfer, testimonials, and card stacking. Respectively, targeting institutions, Trump is 5.9 times more likely to be granting a testimonial (OR = 5.93, p < 0.001), 3.36 times more likely to stack the cards (OR = 3.36, p < 0.001), 2.35 times more likely to use glittering generalities (OR = 2.35, p < 0.001), and 2.03 times more likely to transfer some kind of concept to the image of an institution (OR = 2.03, p < 0.001).

Name calling has a significant negative effect. Trump was 35% less likely to be targeting an institution when name calling (OR = 0.65, p = 0.03). Plain folk and bandwagon were not found to be statistically significant in their usage, although they both involved negative relationships.

4.1.5. Immigrants

Trump used many tactics against immigrants. The strongest positive predictors of Trump’s posts targeting immigrants are glittering generalities, transfer, testimonials, and card stacking, as evidenced in

Table A5. Logistic regression predicting mentions of immigrants.

Predictor	β	SE β	Wald χ2	df	p	OR
(Intercept)	−3.505	0.230	231.68	1	<0.001	0.030
Name Calling	0.055	0.196	0.08	1	0.778	1.057
Glittering Generalities	0.737	0.180	16.73	1	<0.001	2.089
Transfer	1.386	0.188	54.27	1	<0.001	3.999
Testimonials	1.313	0.203	41.89	1	<0.001	3.719
Plain Folk	0.271	0.243	1.25	1	0.263	1.312
Bandwagon	−0.831	0.385	4.65	1	0.031	0.436
Card Stacking	1.114	0.213	27.44	1	<0.001	3.046

. When transferring, it was approximately four times more likely that the target was immigrants (OR = 4.00, p < 0.001); for testimonials, it was 3.72 times more likely (OR = 3.72, p < 0.001); for card stacking, 3.05 times more likely (OR = 3.05, p < 0.001); and finally, Trump using glittering generalities made it 2.09 times more likely that the target was immigrants (OR = 2.09, p < 0.001). Name calling and plain folk appeals both presented positive but not significant results. Bandwagon had a significant negative effect (OR = 0.44, p = 0.03), where posts with bandwagon were 56% less likely to use the tactic against immigrants.

4.1.6. Foreign Leaders

While the results regarding foreign leaders were relatively scarce and non-significant, the strongest predictors of Trump’s posts targeting foreign leaders were testimonials and card stacking. Card stacking made it 2.84 times more likely the post was for or against foreign leaders or actors (OR = 2.84, p = 0.003), while testimonials had a negative relationship with foreign leaders as the target, making it 66% less likely to have them as the target (OR = 0.34, p = 0.02).

4.2. Emotional Appeals and Valence

The second research question posed in this study dealt with how emotional appeals are used in Trump’s online messaging and what the overall valence of that messaging is, illustrated in

Table 3. Correlations between emotion and technique.

Technique	Anger	Fear	Positivity
Name Calling	0.609 ***	0.293 ***	−0.325 ***
Glittering Generalities	−0.179 ***	−0.011	0.622 ***
Transfer	0.069 *	0.176 ***	0.172 ***
Testimonials	−0.368 ***	−0.178 ***	0.277 ***
Plain Folks	0.107 ***	0.169 ***	0.120 ***
Bandwagon	0.004	0.022	0.135 ***
Card Stacking	0.536 ***	0.314 ***	−0.122 ***

* p < 0.05, *** p < 0.001.

. To test the corresponding hypotheses, a series of logit models were used.

The data are consistent with hypothesis 2a. Posts containing propaganda of any kind were approximately 26 times more likely to include an emotional appeal of some kind (OR = 26.19, p < 0.001). Posts containing propaganda were 27 times more likely to contain appeals to anger (OR = 27.62, p < 0.001), 12 times more likely to contain appeals to fear (OR = 12.74, p < 0.001), and 3 times more likely to make appeals to positive emotions such as patriotism, victory, etc. (OR = 3.56, p < 0.001). Applying each of these emotional appeals to every one of the seven propaganda tactics is outside the scope of this study.

Next, the data are also consistent with hypothesis 2b, which supposed that Trump’s propagandistic messaging would be more likely to contain negative than positive or neutral valence. To determine this, multinomial logit models were used to find predicted probabilities. When posts did not contain propagandistic language, the data showed a 6.2% probability that posts were negative, 46.5% neutral, and 47.3% positive. However, posts that did contain propaganda tactics were far more likely (61.6% likelihood) to be negative, with only a 5.4% probability of neutrality and a 33% chance of having a positive valence.

The correlations between each technique and the three types of specific appeals—anger, fear, and positive emotions such as joy, pride, and patriotism—were studied. Anger was strongly positively correlated with name calling and card stacking; it had a negative moderate correlation with testimonials, and weaker correlations with glittering generalities and transfer. Fear appeals had moderate positive correlations with name calling and card stacking, while having weaker correlations with transfer, testimonials, and plain folks. Positive appeals had a strong positive correlation with glittering generalities, a moderate negative correlation with name calling and testimonials, and weaker correlation with transfer, plain folks, bandwagon, and card stacking. Fear in glittering generalities as well as both anger and fear in bandwagon propaganda were not found to be statistically significant.

5. Discussion

This study provides insights into Donald Trump’s online messaging, demonstrating the emotional appeals he uses, who he targets, and how he does so. The findings show Donald Trump relies on propagandistic messaging with emotional appeals and the distinction between “the people” and “elites.” This is aligned with past scholarship, as detailed in the literature review, which explains the ideational approach to populism (Mudde & Kaltwasser, 2017; Rahn, 2019) and defines it as someone who relies on consistent emotional, propagandistic narratives for the people and against elites (Rahn, 2019) Therefore, this study proves Donald Trump is a populist communicator, as hypothesized.

A key contribution of the present work is that it showcases propaganda as something other than a monolithic populist communication strategy used in the same way for every targeted group. Instead, Donald Trump’s messaging as a populist communicator reflects strategic patterns of usage, with different techniques leveraged and deployed against different in- and out-group targets. This includes techniques considered negative being disproportionately used against political opponents, while “positive” techniques (such as glittering generalities, testimonials, and bandwagoning) are disproportionately used for ingroup allies and supporters. These findings suggest propaganda works as a potential persuasive tool and communication strategy for forming narratives about target groups. This study serves as a foundation for similar work comparing the communication styles of populist communicators. This study benefits future work by discussing the causal, persuasive effects of populist messaging.

5.1. Attacking Outgroups and (Sometimes) Defending Ingroups

This study shows Donald Trump does not use the same propaganda techniques to uniformly portray each group. For example, he tends to use ad hominem attacks and name calling against Democrats and the “fake news” media but transfers bad feelings and associations (such as crimes and danger) to immigrants.

Donald Trump attacks outgroups while defending and praising ingroups of all kinds. He appears to perceive Democrats and liberals as the main enemies and a threat to his power. These people and party leaders, such as the previous administration, are painted as the corrupt, powerful elite who pose a threat to the common people, as is typically described by populists. He portrays them as villains, thieves, and criminals rather than legitimate opponents on the political stage.

While Democrats are the enemy elite, ingroup Republicans are considered, in populist parlance, the people. His famous use of the glittering generality “MAKE AMERICA GREAT AGAIN!” seen in many of his posts accompanies testimonials of candidates supposedly saving the country from outside invaders. He frames his own people as the patriotic plain folk returning to and/or helping restore a vague, mythic golden age.

A social media post he repeatedly reuses as an endorsement is a copied and pasted format where he changes few details. This “CTRL + V” type of post is seen in nearly every testimonial or endorsement and contains multiple propaganda tactics—much like the one at the beginning of this paper. An example reads as follows:

“[Name] is a(n) [adjective] America First Candidate running to represent the [positive adjective] People of [location]! As a very successful [past job], [name] knows how to Fight Inflation, Grow the Economy, Lower Taxes, Cut Regulations, and Eliminate Government Waste [add/subtract achievements]. In Congress, [name] will work tirelessly to Secure our Border, Stop Migrant Crime, Strengthen our Incredible Military/Vets, Restore American Energy DOMINANCE, and Defend our always under siege Second Amendment. [add/subtract achievements]. [Name] has my Complete and Total Endorsement—[pronoun] WILL NOT LET YOU DOWN!”

A particularly interesting finding regarding Republicans is that Trump mentions both ingroup followers and supporters as well as outgroup perceived traitors, or RINOs. However, what may be more notable is, according to the empirical findings of this study, he is not likely to appeal to plain folks when mentioning his own party. This is of interest because populist leaders are often seen as one of the people facing a corrupt elite. Donald Trump rarely portrays himself as one of the common people or plain folks.

No discussion on Trump’s rhetoric would be complete without mentioning his claim of “FAKE NEWS!” against legitimate media outlets with proven factual reporting and low levels of bias. He portrays the media as the corrupt enemy of the people, democracy, and himself. He also uses the name-calling technique in this labeling of the “fake news” media, allowing him to use the media as a scapegoat and deflect information that is undesirable to him by dismissing it—as populists are historically known to do. This is an important finding, as this sways his voting base to not partake in the mainstream media. This necessitates further study into countermedia attendance or alt-tech use.

Immigrants were another target of Trump’s emotional tirades and far-right populist propagandistic messaging. The main tactic seen in these attacks was transferring the negative associations of crime and danger to immigrant groups. For example, Trump associated peaceful, law-abiding, undocumented immigrants with murderers, rapists, drug traffickers, gang members, and other dangerous groups, almost always with appeals to fear and anger. Some posts gave graphic descriptions of immigrants murdering children followed by promises to fix the border he claimed was broken by “Border Czar Lyin’ Kamala.” This supports the theory that populist messaging emphasizes excluding an outgroup from “the people.” This can be a dangerous rhetoric tactic, because it could put the targets of these transfers into real danger.

The final group he mentions, foreign leaders and their governments, received the least, though still a significant amount of, attention. He tended to target them through card stacking, rewriting history about the relationships between the American and respective foreign governments, and accepting or giving endorsements to or from other countries, including ones often seen as adversaries of the United States.

5.2. Emotional Appeals and Negativity

This study describes Donald Trump’s personal “brand” as overwhelmingly negative. As Trump is experienced in business, it is possible he is aware of and utilizing this brand intentionally. His communication is nearly twice as likely to be negative than positive, and his propaganda more often than not contains emotional messaging. Even when his posts praise his ingroup followers and supporters, they typically attack someone in an outgroup. In the sample studied, there is a near total lack of simple, positive posts. This is a notable finding, as it further demonstrates the overall message Donald Trump relies on is emotional, negative, and frequently attacks people. Almost nothing can be considered neutral in Trump’s language, suggesting an awareness of widening political polarization and an active attempt to increase it. This is potentially evident in his use of appeals to anger and fear as he rallies his supporters against a corrupt elite enemy. As negativity and a reliance on emotional appeals are a consistent feature of Trump’s messaging, this study’s second hypotheses are supported.

More specifically, Donald Trump’s appeals to anger and fear were most prevalent in instances of name calling and card stacking. Meanwhile, positive appeals to joy, pride, or patriotism were found most often with the use of glittering generalities and testimonials but had a negative relationship with card stacking. This indicates Trump’s usage of card stacking is more likely to focus on the negative and omit the positive than vice versa.

6. Conclusions

This paper sought to provide insights into the propagandistic nature of Donald Trump’s online messaging during his presidential campaign. This was done by describing the seven propaganda tactics and how he used them to target six inductively coded targets based in prior scholarship on propaganda and populism. These variables were analyzed with a series of logit models and found many statistically significant relationships. This study acts as a descriptive example of Trump’s techniques and targets. A limitation of this study is that it may not be generalizable to other populist communicators online. Future research may expand on these findings.

The current study offers several contributions to the literature both on propaganda and populism by providing empirical evidence of how a populist communicator utilizes propaganda techniques. It also demonstrates how the classic propaganda techniques remain relevant on modern social media platforms and in digital environments. Furthermore, it shows empirical evidence of Trump’s usage of these propaganda techniques specifically, providing information for other researchers conducting studies on his campaign messaging, his use of propaganda, and the groups he targets. Much digital social media research focuses on messaging on Twitter/X, Facebook, etc., but this study makes use of Truth Social, an understudied platform.

In conclusion, this study showed empirical evidence that Donald Trump’s messaging during his 2024 presidential campaign is overwhelmingly propagandistic and utilizes all of the traditional propaganda techniques to target his political enemies while supporting his allies.